J. Phys. Chem. A 2000, 104, 4421-4428
4421
Microwave Spectrum, Conformational Equilibrium, Intramolecular Hydrogen Bonding,
Tunneling, and Quantum Chemical Calculations for 1-Ethenylcyclopropan-1-ol
Andrei Leonov,† Karl-Magnus Marstokk,‡ Armin de Meijere,†,§ and Harald Møllendal*,‡,⊥
Department of Chemistry, The UniVersity of Oslo, P.O. Box 1033 Blindern, N-0315 Oslo, Norway, and Institut
für Organische Chemie der Georg-August-UniVersität, Tammannstrasse 2, D-37077 Göttingen, Germany
ReceiVed: NoVember 16, 1999; In Final Form: February 17, 2000
The microwave spectra of 1-ethenylcyclopropan-1-ol, (CH2)2C(OH)CdCH2, and one deuterated species,
(CH2)2C(OD)CdCH2, have been investigated in the 11.0-60.0 GHz region. The (ac,ap)- and (ac,sc1)conformers denoted Syn 1 and Skew 1 were assigned. Each of these two forms is stabilized with an
intramolecular hydrogen bond formed between the hydrogen atom of the hydroxyl group and the π electrons
of the double bond. In the Syn 1 rotamer the CdC-C-O chain of atoms takes a syn conformation (dihedral
angle ) -2.6°) and the H-O-C-Cd link of atoms is gauche (dihedral angle ) -67.2° from syn). The
CdC-C-O link of atoms takes a skew conformation (dihedral angle ) 132.1° from syn) in the Skew 1
rotamer, while the H-O-C-C) dihedral angle is -67.1°. Syn 1 is preferred by 4.9(6) kJ mol-1 relative to
Skew 1. Syn 1 is virtually a hybrid of the most stable conformer of unsubstituted ethenylcyclopropane, and
unsubstituted cyclopropanol, Skew 1, is the corresponding hybrid of the second rotamer of ethenylcyclopropane
and the most stable one of cyclopropanol. The spectrum of Syn 1 is perturbed by tunneling of the hydroxyl
group. An analysis yielded 2280.184(60) MHz for the tunneling frequency and 39.82(19) MHz for the Coriolis
coupling term µca for the normal species. The corresponding values were 72.401(27) and 5.2(10) MHz,
respectively, for the deuterated species. A potential function for the tunneling motion consisting of three
cosine terms was found to have the following potential constants: V1 ) -918.2, V2 ) -900.0, and V3 )
418.0 cm-1. This double-minimum function yields a barrier of 16.6(50) kJ mol-1 at the anti position and
10.6(30) kJ mol-1 at syn. The microwave work has been assisted by ab initio computations at the MP2/ccpVTZ level of theory as well as density functional theory calculations at the B3LYP/6-31G* level. These
calculations indicate that there are only three stable rotameric forms of the molecule. The gas-phase IR spectrum
in the O-H stretching region revealed a broad and complex band red-shifted by roughly 50 cm-1 presumably
as a result of internal hydrogen bonding.
Introduction
Several allylic alcohols that contain the CdC-C-O-H chain
of atoms have been investigated in recent years by microwave
(MW), infrared (IR), and nuclear magnetic resonance (NMR)
spectroscopy as well as by electron diffraction (ED) and
quantum chemical computations.1-7 The molecules studied so
far include the prototype compound 2-propen-1-ol (H2CdCHCH2OH, allyl alcohol)1 and derivatives such as 3-buten-2-ol
(H2CdCHCH(OH)CH3),2 2-methyl-2-propen-1-ol (H2CdC(CH3)CH2OH),3 trans-4a,b and cis-2-buten-1-ol4c (H3CCHdCHCH2OH), 2,3-butadien-1-ol (H2CdCdCHCH2OH),5 1,4-pentadien-3-ol (H2CdCHCH(OH)CHdCH2),6 and 2-cyclopropylideneethanol ((CH2)2CdCHCH2OH).7 Two aromatic compounds, viz. 2-8 and 3-furanmethanol,9 have also been studied.
They are in a sense allylic alcohols because they carry the
CdC-C-O-H link of atoms.
In all these compounds1-9 the preferred rotamer has a heavyatom skew conformation with a CdC-C-O dihedral angle
roughly 120° from syn (0°). The C-C-O-H dihedral angle is
gauche (approximately 60° from syn) in all molecules but
* To whom correspondence should be addressed.
† Institut für Organische Chemie der Georg-August-Universität.
‡ Department of Chemistry, The University of Oslo.
§ E-mail: ameijere1@uni-goettingen.de.
⊥ E-mail: harald.mollendal@kjemi.uio.no.
2-furanmethanol8 (where hydrogen (H) bonding with the oxygen
atom of the ring predominates), allowing a H bond to be formed
with the π electrons of the double bond. A second CdC-C-O
syn rotamer has been found in some of these cases.1j,2a,3b The
conformation of C-C-O-H is gauche in the syn rotamers
too,1c,j,2a,3b allowing a H bond to be formed with the π electrons.
The syn forms are slightly less stable (1-3 kJ mol-1) than their
skew counterparts.1c,j,2a,3b In cases where there are two mirrorimage forms of syn, tunneling of the hydroxyl group is a
prominent feature of the MW spectrum.1j,3a
Formally, 1-ethenylcyclopropan-1-ol (ECP) can be regarded
as an allylic alcohol with the two methylene groups of the
cyclopropyl ring seen as substituents. However, the presence
of this ring is a complicating factor that is likely to have
interesting conformational consequences. In cyclopropane derivatives with unsaturated substituents such as CHO,10 CFO,11
CClO,12 CO2H,13 vinyl,14 or nitro,15 the most stable conformations are those with the plane of the substituent in the symmetry
plane of the molecule. CdC-C-O syn form(s) are thus
expected for ECP when this compound is regarded as a
cyclopropyl derivative with an unsaturated substituent.
Several interesting questions then arise for ECP: e.g., will
the tendency of allyl alcohols to prefer a skew form or the
propensity of cyclopropane derivatives to prefer a syn form
predominate? Are there several rotamers within a narrow energy
10.1021/jp9940755 CCC: $19.00 © 2000 American Chemical Society
Published on Web 04/14/2000
4422 J. Phys. Chem. A, Vol. 104, No. 19, 2000
Leonov et al.
No investigations of the conformational properties of ECP
have been reported. It was therefore decided to carry out a MW
study assisted by quantum chemical calculations and infrared
spectroscopy. MW spectroscopy is ideal for investigating
conformational properties in cases where polar forms are present
because of its high specificity. All conceivable rotamers of the
ECP would each possess a sizable dipole moment that is a
prerequisite for a MW spectrum. Advanced quantum chemical
computations are often found to be useful in predicting rotational
constants, dipole moments, and energy differences for the
various conformers that are sufficiently close to the experimental
ones to be really helpful starting points in the spectral analysis.
In addition, they may give important information about molecular parameters that are not readily accessible experimentally.
Such calculations are therefore of interest in their own right.
Experimental Section
Figure 1. Five possible conformations of 1-ethenylcyclopropan-1-ol.
Only three of these, viz., Syn 1 (ac,ap), Skew 1 (ac,sc1), and Skew 3
(ac,sc2) were found to be “stable” (i.e., minima on the potential energy
hypersurface). Syn 1 and Skew 1 were assigned in this work. Syn 1 is
4.9(6) kJ mol-1 more stable than Skew 1. In the IUPAC nomenclature,
the first denominator stands for the conformation of the cyclopropanol
subunit with respect to the O-H bond and the plane bisecting the
cyclopropane ring, the second for that of the ethenylcyclopropane
subunit.
range? In case this is so, what do they look like? Will any of
them display unusual dynamics, i.e., tunneling? What is the role
of internal H bonding? These are some of the questions we
attempt to answer in this work. It should be added that the
present work is a continuation of the studies of intramolecular
H bonding made in Oslo.16
The conformational properties of ECP are dictated by the
rotation around the O1-C1 and C1-C2 bonds (see Figure 1).
A total of nine all-staggered conformations can thus be
envisaged. Symmetry reduces this number to five that in
principle can be distinguished by spectroscopy provided they
are “stable”, i.e., minima on the potential energy hypersurface.
Representatives of these five forms are shown in Figure 1. The
O1-C1-C2dC3 chain of atoms form a dihedral angle φ of
about +120° measured from syn φ ) 0° in the three skew
conformations shown in Figure 1, The H1-O1-C1-C2
dihedral angle is near -60° (Skew 1), 180° (Skew 2), and +60°
(Skew 3). In the two syn forms, the O1-C1-C2dC3 dihedral
angle is approximately 0°, while the H1-O1-C1-C2 dihedral
angle is about -60° (Syn 1) and 180° (Syn 2). Skew 1 and Syn
1 are both stabilized by a weak intramolecular H bond formed
between the hydroxyl group H atom and the π-electrons of the
C2dC3 double bond. This possibility is absent in the three other
forms.
The sample of 1-ethenylcyclopropan-1-ol utilized in this work
was synthesized according to the published procedure.17 The
compound was purified by low temperature crystallization from
n-pentane, subsequent low temperature zone melting, and finally
distillation. Purity was assessed by 1H and 13C NMR spectroscopy.
The deuterated species was produced by seasoning the
waveguide with D2O and then introducing the parent species.
Roughly 50% deuteration was achieved in this manner. The MW
spectrum was studied using the Oslo Stark spectrometer
described in ref 18. The 11-40 GHz spectral region was
investigated extensively. Selected measurements were also made
in the 40-60 GHz interval. Radio frequency microwave
frequency double resonance (RFMWDR) experiments were
made as described in ref 19 using the equipment mentioned in
ref 20. The microwave absorption cell was cooled to about -25
°C during the experiments. Lower temperatures, which would
have increased the MW spectral intensities, could not be
employed owing to insufficient vapor pressure of the compound.
The pressure was 4-10 Pa when the spectra were recorded and
stored electronically using the computer programs written by
Waal.21 The accuracy of the spectral measurements is presumed
to be better than (0.10 MHz. The resolution is approximately
0.4 MHz. The IR spectrum was taken at room temperature using
a Bruker IFS 88 spectrometer with MCT detection and a long
path minicell from Infrared Analysis. The path length was about
3 m and the pressure was approximately 200 Pa.
Results and Discussion
Quantum Chemical Calculations. The Gaussian 94 program
package22 running on the IBM RS6000 cluster in Oslo was
employed in all the quantum chemical calculations. The 6-31G*
basis set22 and the correlation-consistent polarized triple-ζ basis
set,23 cc-pVTZ, provided with the program were used. Two
different computational schemes, the elaborate MP2/cc-pVTZ
procedure and the much less demanding B3LYP/6-31G*
method, were employed in the computations. In the first of these
procedures, electron correlation was included using the secondorder Møller-Plesset (MP2) perturbation theory.24 In the second
procedure, density functional theory (DFT) calculations were
carried out employing the B3LYP procedure.25 Full geometry
optimization was made in both the MP2 and in the B3LYP
computations.
The reason for selecting these two computational schemes is
that the MP2/cc-pVTZ procedure is assumed to produce accurate
structures26 for the conformations in question. Good estimates
Microwave Spectra of 1-Ethenylcyclopropan-1-ol
of the rotational constants are therefore expected from such
calculations. This of course facilitates the spectral assignments.
It is assumed that the B3LYP/6-31G* structures, dipole
moments, and vibrational frequencies25 are also quite accurate.
Calculations of the structures and vibrational frequencies were
first made with the B3LYP procedure allowing for full geometry
optimization. The B3LYP structures thus obtained were then
used as input in the MP2 computations. MP2 vibrational
frequencies were not computed owing to lack of resources.
Of the five possible conformations in Figure 1, only three,
viz. Syn 1 as well as Skew 1 and Skew 3 were found to be
stable in the B3LYP calculations, because no negative vibrational frequencies27 were computed for any of them. No stable
Syn 2 or Skew 2 forms were identified in these calculations.
The B3LYP computations in fact refined to one of the three
stable forms mentioned above in each case when conformations
corresponding to Syn 2 or Skew 2 were used as starting point
in the computations. Because the MP2/cc-pVTZ calculations
are so comprehensive, no calculations for an assumed starting
geometry of the hypothetical Syn 2 or Skew 2 forms were
carried out using this computational scheme. It is believed on
the basis of the B3LYP calculations that Syn 2 and Skew 2
indeed do not exist as stable forms of ECP, and that the other
three conformers (Syn 1, Skew 1, and Skew 2) are the only
stable forms of this compound. The hypothetical Syn 2 and
Skew 2 conformations have an anti conformation for the H1O1-C1dC2 chain of atoms. It is interesting to note that only
the gauche form has been found for the related compound
cyclopropanol in a MW study.28 It thus appears that the H1O1-C1dC2 anti conformation is generally not stable in
cyclopropanols.
The MP2 geometries of the three stable conformers are given
in Table 1 together with some other parameters of interest. The
corresponding parameters obtained in the B3LYP calculations
are rather similar to these. They are given in Table 1S in the
Supporting Information.
There is nothing unusual about the bond lengths, bond angles,
and dihedral angles of the three stable rotamers (Table 1).
Interestingly, Syn 1 is predicted to be stabilized by more than
7 kJ mol-1 relative to the other two stable forms (Skew 1 and
Skew 3). The tendency to prefer syn forms with respect to the
ethenyl moiety for cyclopropyl derivatives with unsaturated
groups thus seems to be of overriding importance (see above).
The internal H bond interaction is optimal in both Syn 1 and
Skew 1. It is assumed that this interaction is the reason Skew
1 is preferred to Skew 3, which cannot have a H bond. The
MP2 distances between the H1 and C2 and C3 atoms calculated
form the structure in Table 1 are 266 and 286 pm, respectively,
in Syn 1 and 263 and 356 pm in Skew 1. The sum of the van
der Waals distances for hydrogen (120 pm29) and aromatic
carbon (170 pm29) is 290 pm and thus slightly longer than the
distances between H1 atom and C2 atom in these two cases.
This indicates a modest H bond strength.
Another indication that the H bond interaction must be rather
weak, but definitely not absent, comes from the IR spectrum
of the O-H stretching frequency of the gas. It is seen in Figure
2 that this rather broad band has peaks at 3631 and 3625 cm-1
and a shoulder at 3620 cm-1. The O-H stretching frequency
of methanol30 falls at 3682 cm-1. The O-H stretching frequency
of the title compound is thus red-shifted by roughly 50 cm-1
as compared to methanol.
Syn 1 has a mirror-image form (not distinguishable by MW
spectroscopy) that can be produced by rotating the H1-O1C1dC2 dihedral angle until it becomes +67.2° instead of
J. Phys. Chem. A, Vol. 104, No. 19, 2000 4423
TABLE 1: Structure,a Rotational Constants, Dipole
Moments, and Energy Differences of Syn 1, Skew 1, and
Skew 3 of 1-Ethenylcyclopropan-1-ol As Calculated at the
MP2/cc-pVTZ Level of Theory
conformer
Skew 1
Skew 3
Bond Length/pm
96.4
140.0
147.3
133.6
108.4
107.9
108.0
149.9
151.1
107.9
107.9
107.9
108.0
Syn 1
96.4
140.9
148.0
133.5
108.5
108.0
108.1
149.6
150.1
107.9
107.9
107.9
107.9
96.3
140.9
148.1
133.5
108.6
108.0
108.0
151.0
148.9
107.9
108.0
107.8
107.8
Angle/deg
106.6
116.5
124.2
115.8
120.9
120.8
113.9
117.1
117.9
115.2
115.7
118.2
106.5
114.0
125.6
114.6
120.9
121.5
112.9
117.0
117.6
115.8
115.8
119.3
107.1
114.6
126.3
114.8
120.8
121.7
116.4
113.7
117.6
116.5
115.2
119.0
Dihedral Angleb/deg
-67.2
-67.1
-2.6
132.1
176.9
-47.6
179.4
179.0
-1.8
-0.8
148.5
151.6
81.1
83.9
142.5
143.2
-0.7
-1.3
-3.9
-4.7
-148.4
-150.6
72.9
135.3
-45.7
178.0
-2.2
-72.7
-140.4
148.9
3.3
0.2
-144.2
A
B
C
Rotational Constants/MHz
6431.8
5505.7
2964.4
3180.7
2435.2
2517.8
5507.0
3192.7
2506.3
µa
µb
µc
µtot
Dipole Momentc/10-30 C m
0.13
3.34
2.80
1.83
4.07
2.94
4.94
4.80
0.97
0.97
5.13
5.30
Energy Differenced/kJ mol-1
0.0e
6.97
12.35
O1-H1
C1-O1
C1-C2
C2-C3
C2-H2
C3-H3
C3-H4
C1-H4
C1-C5
C4-H5
C4-H6
C5-H7
C5-H8
H1-O1-C1
O1-C1-C2
C1-C2-C3
C1-C2-H2
C2-C3-H3
C2-C3-H4
O1-C1-C4
O1-C1-C5
C1-C4-H5
C1-C4-H6
C1-C5-H7
C1-C5-H8
H1-O1-C1-C2
O1-C1-C2-C3
O1-C1-C2-H2
C1-C2-C3-H3
C1-C2-C3-H4
H1-O1-C1-C4
H1-O1-C1-C5
O1-C1-C4-H5
O1-C1-C4-H6
O1-C1-C5-H7
O1-C1-C5-H8
Atom numbering is given in Figure 1. Measured from syn ) 0°.
1 debye ) 3.335 64 × 10-30 C m. d Relative to Syn 1. e Total energy
obtained in the MP2/cc-pVTZ computations: -708 784.40 kJ mol-1.
a
b
c
-67.2° as for the conformer shown in Figure 1. A doubleminimum potential will be associated with this rotation around
the O1-C1 bond. The presence of this double-minimum
potential manifests itself in the MW spectrum (see below).
B3LYP calculations were made at intervals of 30° of the said
dihedral angle allowing all other structural parameters to vary
freely. The resulting potential function was predicted to have a
maximum of 14.4 at the syn (0°) and 15.2 kJ mol-1 at the anti
(180°) position of the H1-O1-C1dC2 dihedral angle. The
existence of no further conformations was indicated in these
computations.
4424 J. Phys. Chem. A, Vol. 104, No. 19, 2000
Figure 2. Gas-phase infrared spectrum of 1-ethenylcyclopropan-1-ol
in the O-H stretching region. This band has maxima at 3631 and 3625
cm-1 and a shoulder at 3620 cm-1.
MW Spectrum and Assignment of Syn 1. The survey
spectra taken at field strengths of about 100 and 1000 V cm-1
revealed a very dense, medium intense spectrum with absorption
lines occurring every few megahertz throughout the entire
spectral range. Some unusual features were immediately seen
in the 20-22 and 15-16 GHz regions where numerous
comparatively strong transitions with rapid Stark effects coalesce
in band heads at 21 375 and 21 706 MHz. Similar, but less
intense spectral lines coalesce at 15 334, 15 618, and 15 736
Leonov et al.
MHz. The first mentioned band heads are shown in Figure 3. It
is assumed that all five band heads somehow are associated
with the tunneling of the hydroxyl group, but we were unable
to give definite assignments.
The ab initio results in Table 1 predict that Syn 1 is the most
stable conformer. Its largest dipole moment components (Table
1) are predicted to lie along the b- and c-inertial axes. As
tunneling was suspected for this rotamer, searches were first
made for b-type transitions since they were assumed to deviate
less from “normal” behavior than the c-type transitions do.
Searches for the strong medium-J b-type Q-branch transitions
of Syn 1 were first made. These lines, which are among the
strongest ones in the spectrum, were soon identified. They
always appeared as doublets of equal intensity. The separation
between the doublets varied from not being resolved (less than
about 0.4 MHz) up to about 60 MHz at most.
The strongest low-J bR-branch lines were searched for next,
and again found as doublets. These b-type transitions were first
fitted to Watson’s Hamiltonian31 including quartic centrifugal
distortion constants. The root-mean-square (rms) deviations of
these fits were roughly 1 MHz, 10 times the experimental
uncertainty. The frequencies of strong c-type transitions were
now predicted using this Hamiltonian. No candidates for these
c-type lines could be found in the vicinity of the predicted
frequencies. It was then clear that a more sophisticated approach
had to be followed in order to assign the c-type lines.
The unusual spectrum of Syn 1 is explained as follows: The
H atom of the hydroxyl group performs a large-amplitude
motion governed by a double-minimum potential associated with
the torsion around the C1-O1 bond. The spatial direction of
the c-component of the dipole moment is inverted when one
conformation is transformed into its mirror image. The ground
state is a symmetrical or (+) state denoted 0+. The first excited
state is an antisymmetrical or (-) state denoted 0-. The energy
separation between these two states is denoted ∆, often called
the tunneling frequency. The selection rules for the a- and b-type
Figure 3. MW spectrum in the 21-22 GHz spectral region showing the pileups at 21 375 and 21 706 MHz. The intensity scale is arbitrary. The
spectrum was taken at a field strength of about 100 V cm-1.
Microwave Spectra of 1-Ethenylcyclopropan-1-ol
J. Phys. Chem. A, Vol. 104, No. 19, 2000 4425
TABLE 2: Spectroscopic Constantsa,b of the 0+ and 0- States of the Syn 1 Conformer of 1-Ethenylcyclopropan-1-ol
normal species
0+
A/MHz
B/MHz
C/MHz
∆c/MHz
µca/MHz
∆J/kHz
∆JK/kHz
∆K/kHz
δJ/kHz
δK/kHz
no. of transitions in fit
rmsd deviation/MHz
maximum value of J
deuterated species
0-
6385.4056(84)
6385.3017(96)
2937.5418(61)
2937.5384(62)
2415.0811(69)
2415.0792(72)
2280.184(60)
39.82(19)
0.286(30)
0.254(30)
1.561(31)
2.624(43)
5.13(24)
-3.17(35)
0.0494(10)
0.0643(13)
-0.135(25)
-0.383(33)
324
0.244
66
0+
0-
6104.3616(50)
2926.0885(27)
2378.9124(40)
6104.3449(54)
2926.0655(27)
2378.9104(40)
72.401(27)
5.2(10)
0.385(13)
2.004(32)
1.08(13)
0.0470(18)
0.144(50)
0.355(12)
2.319(29)
-0.11(12)
0.0616(17)
-0.247(47)
160
0.096
70
a
As defined by Nielsen.32 b Uncertainties represent one standard deviation. c ∆ ) W01 (energy separation between 0- and 0+ states32). d Rootmean-square.
transitions of the 0+ and 0- states follow rigid-rotor selection
rules. The selection rules for the c-type transitions are those of
a rigid rotor plus (-) r (+), or (-) r (+). In this case large
deviation from rigid-rotor behavior is expected. A splitting of
approximately 2∆ is thus expected for each c-type rotational
transition.
Fortunately, a computer program has been written by
Nielsen32 to deal with spectra of this type. This program is based
on the reduced Hamiltonian defined as follows:32
Hred ) |0⟩ {Hr(0) + Hd(0)}⟨0| + |1⟩{Hr(1) + Hd(1) + W01}
⟨1| + |0⟩Hc⟨1| + |1⟩Hc⟨0|
where the label 0 corresponds to the (+) state and the label 1
corresponds to the (-) state.
Using Ir representation,31 one has
Hr(ν) ) B(ν)Jb2 + C(ν)Jc2 + A(ν)Ja2
where ν refers to the (+) or the (-) state, respectively.
Hd ) {Watson31 quartic and sextic centrifugal
distortion constants}(ν)
W01 ) ⟨1|Hvib0|1⟩ - ⟨0|Hvib0|0⟩
W01 corresponds to ∆ above.
Hc is the Coriolis term of the form
Hc ) µij⟨JjJi + JiJj⟩
where i and j refer to the principal inertial axis. µij is the reduced
moment of inertia.
In the first attempts to find the c-type transitions ∆ was
assumed to be about 21 375 MHz, which is the most intense of
the band heads. The coalescing lines were assumed to be high-J
c-type Q-branch transitions. However, no successful assignments
could be made in this manner. The alternative band head
frequencies of 21 706, 15 334, 15 618, and 15 736 MHz were
then used with the same negative result.
It was then decided to see if we could be more successful
with the spectrum of the deuterated species, (CH2)2C(OD)CdCH2.
Tunneling should be reduced by at least 1 order of magnitude
in this isotopomer as compared to the normal species, making
assignments easier. The assignment of the b-type transitions was
straightforward. Nearly all of these doublet lines had now
coalesced into a single line. Only a few of them were split by
up to about 0.7 MHz. The frequencies of the c-type transitions
were then predicted from these b-type lines. Pairs of lines split
by roughly 100-150 MHz were now readily assigned. Their
average frequencies agreed well with the frequencies predicted
from the b-type transitions.
Nielsen’s program was now employed. Only quartic centrifugal distortion constants were included because centrifugal
distortion is not very prominent in this compound. Various
Coriolis coupling schemes were tested. It turned out that fits
using µca was the most efficient ones. In fact, µca was the only
coupling constant for which a significant value could be
obtained. A total of 160 transitions with a maximum value of
J ) 70 were used to derive the spectroscopic constants listed
in Table 2. Only the resolved b-type doublets were used in this
fit. It is seen from this table that ∆ ) 72.400(27) and µca )
5.2(10) MHz, and the fit itself is as good as the experimental
uncertainty with a rms deviation ) 0.096 MHz. A full listing
of the observed transitions and the least-squares fit residuals is
given in Table 3S in the Supporting Information.
We now turned to the normal species again. From a similar
compound, cyclopropanol,28 it is known that the tunneling
frequency ∆ of the normal species is about 25 times as large as
the corresponding ∆ of the deuterated species. A value of about
2000 MHz was thus guessed for the normal species of ECP.
With this in mind, a few trial fits resulted in the assignment of
the c-type lines. The fitting procedure was then carried out in
the same manner as for the deuterated species. The final results
using 324 b- and c-type transitions (given in Table 2S in the
Supporting Information) are presented in Table 2. No a-type
lines were identified presumably because they are too weak
owing to a small µa (Table 1).
The tunneling frequency ∆ ) 2280.184(60) MHz is about
30 times larger than the corresponding value of the deuterated
species, whereas µca ) 39.82(19) MHz is approximately 8 times
larger than its counterpart in the deuterated compound. The rms
deviation of the fit is 0.244 MHz, which is quite satisfactory.
Inclusion of the Coriolis coupling constants µab and µbc in the
least-squares fit yielded no improvement. The standard deviations of these two parameters were found to be about as large
as the constants themselves. µca was therefore the only coupling
constant retained in the final fit.
It should be mentioned that some of the b-type Q-branch
transitions deviate from this fit by up to approximately 3 MHz.
The lines in question are some of those that fall at the lowest
frequencies of the differentb Q-branch series (where the series
“turns around” and moves toward higher frequencies with higher
values of the J quantum number). This deviation is assumed to
4426 J. Phys. Chem. A, Vol. 104, No. 19, 2000
Leonov et al.
TABLE 3: Spectroscopic Constantsa,b of the Ground
Vibrational State of the Skew 1 Conformer of
1-Ethenylcyclopropan-1-ol
A/MHz
B/MHz
C/MHz
∆J/kHz
∆JKc/kHz
no. of transitions in fit
rms deviation/MHz
5506.94(24)
3161.668(19)
2494.889(21)
0.823(61)
7.179(99)
21
0.147
a
A reduction, Ir representation.31 b Uncertainties represent one
standard deviation. c Further quartic constants preset at zero.
Figure 4. Potential function V(φ) ) (1/2)V1(1 - cos φ) +
(1/2)V2(1 - cos 2φ) + (1/2)V3(1 - cos 3φ) with V1 ) -918.2, V2 )
-900.0, and -V3 ) 418.0 cm-1 for the internal rotation of the hydroxyl
group of the Syn 1 (ac,ap)-rotamer of 1-ethenylcyclopropan-1-ol.
arise from higher order effects not accounted for in the
Hamiltonian employed here. These transitions (less than 10)
were omitted in the final fit.
The tunneling frequency of gauche cyclopropanol is 4115.26(42) MHz,28 similar to the value in Table 2 (2280.184(60) MHz).
It is interesting that H bonding with the π electrons does not
seem to influence the tunneling frequency to any great extent.
The ∆ ) 2280.184(60) MHz frequency is much less than
the five pileup frequencies (15 334, 15 618, 15 736, 21 375, and
21 706 MHz) mentioned above. It is suggested that these five
frequencies have something to do with the tunneling motion of
the hydroxyl group, presumably excited states, but no detailed
explanation can be offered. Similar pileups have been observed
before, e.g., in the tunneling spectrum of ethylene glycol (at
about 17.1 GHz), and remain unassigned to this date.33
Potential Function of the Hydroxyl Group Rotation. Some
information that can be used to calculate a potential function
for the torsional vibration of the hydroxyl group is available:
The tunneling frequencies (∆) for the normal and the deuterated
species are known experimentally (Table 2). The H1-O1C1dC2 dihedral angle (henceforth called φ) was assumed to
have the MP2 value shown in Table 1 (67° from syn, 113° from
anti). The syn and anti positions were assumed to be maxima
in agreement with the B3LYP calculations (see above).
A simplified Hamiltonian is assumed for the internal rotation:
H ) Fp2 + V(φ)
where the potential function V(φ) consists of three cosine terms:
V(φ) ) (1/2)V1(1 - cos φ) + (1/2)V2(1 - cos 2φ) +
(1/2)V3(1 - cos 3φ)
The F value was calculated to be 19.706 cm-1 for the normal
species and 9.903 cm-1 for the deuterated species using the
method of Pitzer and Gwinn.34 The eigenvalues of the Hamiltonian were calculated as described by Lewis et al.35 employing
a computer program written by Bjørseth.35 φ was assigned a
value of 0° at the anti position, as is often done. V(φ) was
assumed to have its minimum at 113° from anti (67° from syn).
The potential constants V1, V2, and V3 were then adjusted to
give the best fit to the data. This was obtained with V1 )
-918.2, V2 ) -900.0, and V3 ) 418.0 cm-1. The resulting
double-minimum-potential function is drawn in Figure 4. The
maximum at anti position (φ ) 0°) is calculated to be 1387.2
cm-1 (16.59 kJ mol-1). The maximum at syn (φ ) 180°) is
887.0 cm-1 (10.61 kJ mol-1). The potential function depends
very critically on the assumptions made, and it is difficult to
make estimates of the uncertainties. The best values we can
come up with are 16.6(50) for the anti and 10.6(30) kJ mol-1
for the syn barrier, respectively. These values are rather close
to those found in the B3LYP calculations above (15.2 and 14.4
kJ mol-1, respectively) and similar to those found for cyclopropanol.28
Assignment of Skew 1. This conformer is predicted to be
less stable than Syn 1 by as much as 7.0 kJ mol-1 (Table 1).
The best way to assign such high-energy rotamers with relatively
weak spectra “drowned” in the dense and strong spectrum of
another conformer is to use the RFMWDR technique19 because
it is so specific, provided that the prolate conformer in question
has a sizable dipole moment component along the a-inertial axis.
This was fortunately predicted to be the case for Syn 1 (µa ≈
3.3 × 10-30 C m; Table 1). Model calculations using the
rotational constants in Table 1 indicate a frequency difference
of about 4.5 MHz between the 96,3 and 96,4 energy levels. The
RF pump was set at this frequency. The 106,4 r 96,3 and
106,5 r 96,4 pair of transitions was searched for around 57.3
GHz, while the 96,3 r 86,2 and the 96,4 r 86,3 pair was looked
for in the vicinity of 51.5 GHz. These double resonance searches
met with immediate success. The assignments were then
extended to further aR transitions using RFMWDR spectroscopy
where possible. Additional assignments of high-K-1 lines of the
same kind could be made at low (about 50 V cm-1) field
strengths. Low-K-1 lines could not be assigned with certainty
because they are slowly modulated and often overlapped by
the much stronger transitions of the Syn 1 rotamer present nearly
everywhere in the spectrum. A total of 21 a-type transitions
shown in Table 4S in Supporting Information were assigned.
No b- or c-type lines could be found. The spectroscopic
constants (A reduction Ir representation31) are shown in Table
3. Only two of the quartic centrifugal constants could be
determined from the collection of transitions available. Assignment of the deuterated species was not attempted owing to the
low intensity of the spectrum of this high-energy conformer.
It is seen in Table 1 that Skew 1 and Skew 3 should have
very similar rotational constants. However, the dipole moments
have radically different orientations within the molecule. Skew
1 has its largest component along the a-inertial axis, whereas
Skew 3 has its largest component along the c-axis. The fact
that aR transitions are the only ones assigned is taken as evidence
that Skew 1 has not been confused with Skew 3.
The energy difference between Skew 1 and Syn 1 was
determined by relative intensity measurements made as in the
manner described in ref 37. The energy difference depends on
the ratio of the dipole moments squared.38 This ratio was
calculated from the values of the dipole moments listed in Table
1. The energy difference was found to be 4.9 kJ mol-1. One
standard deviation was estimated to be (0.6 kJ mol-1. The
Microwave Spectra of 1-Ethenylcyclopropan-1-ol
energy difference of 4.9(6) kJ mol-1 is in fair agreement with
the MP2 result (7.0 kJ mol-1).
Skew 3 is predicted to be about 12 kJ mol-1 less stable than
Syn 1 (Table 1) This rotamer is computed to have its major
dipole moment component along the c-inertial axis. Finding such
a high-energy form in this dense spectrum would undoubtedly
be very difficult. Our attempts to do so were futile.
The strongest transitions present in this spectrum have been
assigned to Syn 1. Yet, the above assignments of the spectra of
Syn 1 and Skew 1 represent only a fraction of the very large
number of transitions present. The B3LYP calculations indicate
that there are several low-lying vibrational modes that are well
populated at -25 °C. Each of these states is expected to have
a tunneling spectrum of its own. No assignments of vibrationally
excited states were made.
Structure. The substitution coordinates39 of H1 of Syn 1 were
calculated as |a| ) 31.966(61), |b| ) 174.633(13), and |c| )
78.662(28) pm, respectively, from the rotational constants and
their standard deviations of the 0+ states of the normal and
deuterated species. The actual uncertainties are much larger than
these formal standard deviations (in parentheses) owing to the
large-amplitude motions that the H1 atom performs. The values
calculated from the structure in Table 1 are |a| ) 31.37, |b| )
176.52, and |c| ) 76.38 pm, respectively, and are in good
agreement with the substitution values. This is one indication
that the structure in Table 1 is close to the “real” structure of
this rotamer. Another indication is the fact that the rotational
constants in Tables 1 and 2 deviate by less than 1% in each
case. The same is seen for Skew 1. This agreement is not
believed to be fortuitous, but in fact reflects that MP2/cc-pVTZ
structures are close to the “correct” structures, as has already
been pointed out.26 The structures of Syn 1 and Skew 1 shown
in Table 1 are therefore adopted as plausible structures. It is
expected that any experimental structures that might be determined in the future will be very close to those shown in Table
1.
Conclusions
Two conformers of ECP denoted Syn 1 and Skew 1 have
been assigned using MW spectroscopy. The two forms are
interconverted by rotating approximately 120° around the C1C2 bond. Syn 1 is the most stable form of the molecule, being
4.9(6) kJ mol-1 more stable than Skew 1. The prevalence of
these two conformers is a consequence of several factors: Both
rotamers are stabilized by a weak internal H bond formed
between the H atom of the hydroxyl group and the π electrons
of the ethenyl group. Tunneling of the hydroxyl group is a
prominent feature of the spectrum of Syn 1. A double-minimum
potential could be determined for this species. Interaction
between the lone pairs on oxygen and the electrons of the
cyclopropyl group and the conjugative interaction between the
ethenyl and the cyclopropyl group are likely to also contribute
to the stabilities of these two conformers. Essentially, the
preferred (ac,ap)-rotamer Syn 1 is the hybrid of the most stable
conformatons of the ethenylcyclopropane14 and the cyclopropanol28 subunits, while the (ac,sc1)-rotamer Skew 1 is the hybrid
of the second stable rotamer of unsubstituted ethenylcyclopropane14 and the most stable one of unsubstituted cyclopropanol.28
Ab initio calculations at the MP2/cc-pVTZ level provide reliable
structures and energy differences for the two rotamers.
Acknowledgment. Anne Horn is thanked for the art work
and for experimental assistance. This work has received support
from the Research Council of Norway (Programme for Supercomputing) through a grant of computing time.
J. Phys. Chem. A, Vol. 104, No. 19, 2000 4427
Supporting Information Available: Table 1S contains the
B3LYP structures. Tables 2S-4S contain the MW transitions
used to determine the spectroscopic constants shown in Tables
2 and 3. This material is available free of charge via the Internet
at http://pubs.acs.org.
References and Notes
(1) (a) Murty, A. N.; Curl, R. F. J. Chem. Phys. 1967, 46, 4176. (b)
Badawi, H.; Lorencak, P.; Hillig, K. W.; Imachi, M.; Kuczkowski, R. L. J.
Mol. Struct. 1987, 162, 247. (c) Vanhouteghem, F.; Pyckhout, W.; van
Alsenoy, C.; van den Enden, L.; Giese, H. J. J. Mol. Struct. 1986, 140, 33.
(d) Aspiala, A.; Lotta, T.; Murto, J.; Räsänen, M. J. Chem. Phys. 1983, 79,
4183. (e) Schleyer, P. v. R.; Trifani, D. S.; Bacskai, R. L. J. Am. Chem.
Soc. 1958, 80, 6691. (f) Oki, M.; Iwamura, H. Bull. Chem. Soc. Jpn. 1959,
32, 567. (g) Bakke, J. M.; Schie, A. M.; Skjetne, T. Acta Chem. Scand.
Ser. B 1986, 40, 703. (h) Brouwer, H.; Stothers, J. B. Can. J. Chem. 1972,
50, 1361. (i) Pierre, J.-L. Bull. Soc. Chim. Fr. 1970, 3116. (j) Melandri, S.;
Favero, P.; Caminati, W. Chem. Phys. Lett. 1994, 223, 541. (k) Silvi, B.;
Perchard, J. P. Spectrochim. Acta A 1976, 32, 11.
(2) (a) Smith, Z.; Carballo, N.; Wilson, E. B.; Marstokk, K.-M.;
Møllendal, H. J. Am. Chem. Soc. 1985, 107, 1951. (b) Kahn, S. D.; Hehre,
W. J. Tetrahedron Lett. 1985, 26, 3647. (c) Kahn, S. D.; Hehre, W. J. J.
Am. Chem. Soc. 1987, 109, 666.
(3) (a) Caminati, W.; Fantoni, A. C. Chem. Phys. 1986, 105, 59. (b)
Caminati, W.; Fantoni, A. C.; Velino, B.; Siam, K., Schäfer, L.; Ewbank,
J. D.; van Alsenoy, C. J. Mol. Spectrosc. 1987, 124, 72.
(4) (a) Lum, D. K.; Baumann, L. E.; Malloy, T. B., Jr.; Cook, R. L. J.
Mol. Struct. 1987, 50, 1. (b) Caminati, W.; Velino, B.; Fantoni, A. C. Mol.
Phys. 1987, 61, 1269. (c) Caminati, W.; Paolucci, C.; Velino, B. J. Mol.
Spectrosc. 1989, 137, 362.
(5) Horn, A.; Marstokk, K.-M.; Møllendal, H.; Priebe, H. Acta Chem.
Scand. A 1983, 37, 679.
(6) Marstokk, K.-M.; Møllendal, H. Acta Chem. Scand. 1990, 44, 18.
(7) Bräse, S.; Klæboe, P.; Marstokk, K.-M.; de Meijere, A.; Møllendal,
H.; Nielsen, C. J. Acta Chem. Scand. 1998, 52, 1122.
(8) Marstokk, K.-M.; Møllendal, H. Acta Chem. Scand. 1994, 48, 25.
(9) Marstokk, K.-M.; Møllendal, H. Acta Chem. Scand. 1993, 47, 849.
(10) Volltrauer, H. N.; Schwendeman, R. H. J. Chem. Phys. 1971, 54,
260.
(11) Volltrauer, H. N.; Schwendeman, R. H. J. Chem. Phys. 1971, 54,
268.
(12) Nair, K. P. R.; Boggs, J. E. J. Mol. Struct. 1976, 33, 45.
(13) Marstokk, K.-M.; Møllendal, H.; Samdal, S. Acta Chem. Scand.
1991, 45, 37.
(14) (a) Ethenylcyclopropane exists in the gas phase at ambient
temperature as an equilibrium mixture of ∼75% ap- and ∼25% scconformers: de Meijere, A.; Lüttke, W. Tetrahedron 1969, 25, 2047. (b)
Codding, E. G.; Schwendeman, R. H. J. Mol. Spectrosc. 1974, 49, 226.
(15) Mochel, A. R.; Britt, C. O.; Boggs, J. E. J. Chem. Phys. 1973, 58,
3221.
(16) Recent examples include: (a) 1-amino-1-ethenylcyclopropane,
Marstokk, K.-M.; de Meijere, A.; Møllendal, H.; Wagner-Gillen, K. J. Phys.
Chem. A 2000, 104, 2897. (b) 1-amino-1-ethynylcyclopropane, Marstokk,
K.-M.; de Meijere, A.; Wagner-Gillen, K.; Møllendal, H. J. Mol. Struct.
1999, 509, 1. (c) 3,3,3-trifluoropropanol, Marstokk, K.-M.; Møllendal, H.
Acta Chem. Scand. 1999, 53, 202. (d) dicyclopropyl carbinol, Marstokk,
K.-M.; Møllendal, H. Acta Chem. Scand. 1997, 51, 800. (e) 1,2-diphosphinoethane, Marstokk, K.-M.; Møllendal, H. Acta Chem. Scand. 1997, 51,
875. (f) 1,2-ethanedithiol, Marstokk, K.-M.; Møllendal, H. Acta Chem.
Scand. 1997, 51, 653. (g) 2-nitroethanol, Marstokk, K.-M.; Møllendal, K.M. Acta Chem. Scand. 1996, 50, 505. (h) 1-cyclopropylcyclopropanol,
Chaplinski, V.; Marstokk, K.-M.; de Meijere, A.; Møllendal, H. Acta Chem.
Scand. 1996, 50, 486. See also reviews: (i) Wilson, E. B.; Smith, Z. Acc.
Chem. Res. 1987, 20, 257. (j) Møllendal, H. Structures and Conformations
of Non-Rigid Molecules; Laane, J., Dakkouri, M., van der Veken, B.,
Oberhammer, H., Eds.; Kluwer Academic Publishers: Netherlands, 1993;
p 277.
(17) Wasserman, H. H.; Hearn, M. J.; Cochoy, R. E. J. Org. Chem.
1980, 45, 2874.
(18) Guirgis, G. A.; Marstokk, K.-M.; Møllendal, H. Acta Chem. Scand.
1991, 45, 489.
(19) Wordarczyk, F. J.; Wilson, E. B. J. Mol. Spectrosc. 1971, 37, 445.
(20) Marstokk, K.-M.; Møllendal, H. Acta Chem. Scand. A 1988, 42,
374.
(21) Waal, Ø. Personal communication, 1994.
(22) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.;
Johnson, B. G.; Robb, M. A.; Cheeseman, J. R.; Keith, T.; Petersson, G.
A.; Montgomery, J. A.; Raghavachari, K.; Al-Laham, M. A.; Zakrzewski,
V. G.; Ortiz, J. V.; Foresman, J. B.; Peng, C. Y.; Ayala, P. Y.; Chen, W.;
Wong, M. W.; Andres, J. L.; Replogle, E. S.; Gomperts, R.; Martin, R. L.;
Fox, D. J.; Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart, J. P.; Head-
4428 J. Phys. Chem. A, Vol. 104, No. 19, 2000
Gordon, M.; Gonzalez, C.; Pople, J. A. Gaussian 94, Revision B.3; Gaussian
Inc.: Pittsburgh, PA, 1995.
(23) (a) Dunning, T. H. J. Chem. Phys. 1989, 90, 1007. (b) Kendall, R.
A.; Dunning, T. H.; Harrison, R. J. J. Chem. Phys. 1992, 96, 6796. (c)
Dunning, T. H.; Woon, D. E. J. Chem. Phys. 1993, 98, 1358. (d) Dunning,
T. H.; Woon, D. E. J. Chem. Phys. 1994, 100, 2975. (e) Dunning, T. H.;
Woon, D. E. J. Chem. Phys. 1995, 103, 4572.
(24) Møller, C.; Plesset, M. S. Phys. ReV. 1934, 46, 618.
(25) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(26) Helgaker, T.; Gauss, J.; Jørgensen, P.; Olsen, J. J. Chem. Phys.
1997, 106, 6430.
(27) Hehre, W. J.; Radom, L.; Schleyer, P. v. R.; Pople, J. A. Ab Initio
Molecular Orbital Theory; Wiley: New York, 1985; p 227.
(28) MacDonald, J. N.; Norbury, D.; Sheridan, J. J. Chem. Soc., Faraday
Trans. 2 1978, 74, 1365
(29) Pauling, L. The Nature of the Chemical Bond, 3rd ed.; Cornell
University Press: New York, 1960; p 260.
Leonov et al.
(30) Richter, W.; Schiel, D. Ber. Bunsen-Ges. Phys. Chem. 1981, 85,
548.
(31) Watson, J. K. G. In Vibrational Spectra and Structure; Durig, J.
R., Ed.; Elsevier: Amsterdam, 1977; Vol. 6.
(32) Nielsen, C. J. Acta Chem. Scand. A 1977, 31, 791.
(33) Marstokk, K.-M.; Møllendal, H. J. Mol. Struct. 1974, 22, 301. (b)
Christen, D.; Coudert, L. H.; Suenram, R. D.; Lovas, F. J. J. Mol. Spectrosc.
1995, 172, 57. (c) Christen, D. Personal communication, 1999.
(34) Pitzer, K. S.; Gwinn, W. D. J. Chem. Phys. 1942, 19, 428.
(35) Lewis, J. D.; Malloy, Jr., T. B.; Chao, T. H.; Laane, J. J. Mol.
Struct. 1972, 12, 427.
(36) Bjørseth, A. Personal communication, 1974.
(37) Esbitt, A. S.; Wilson, E. B. ReV. Sci. Instrum. 1963, 34, 901.
(38) (a) Caminati, W. J. Mol. Spectrosc. 1987, 121, 61. (b) Townes, C.
H.; Schawlow, A. L. MicrowaVe Spectroscopy, McGraw-Hill: New York,
1956.
(39) Kraitchman, J. Am. J. Phys. 1953, 21, 17.